Compétences agent-tools
📦

agent-tools

Risque faible ⚙️ Commandes externes🌐 Accès réseau

Run 150+ AI Apps via inference.sh CLI

Également disponible depuis: inference-sh-8,inferencesh,inference-sh-6,inference-sh-3,inf-sh,inference-sh-0

Access cloud-based AI models for image generation, video creation, LLMs, and search without local GPU setup. Execute inference.sh workflows directly through Claude Code using simple CLI commands.

Prend en charge: Claude Codex Code(CC)
🥉 72 Bronze
1

Télécharger le ZIP du skill

2

Importer dans Claude

Allez dans Paramètres → Capacités → Skills → Importer un skill

3

Activez et commencez à utiliser

Tester

Utilisation de "agent-tools". Generate an image of a cat astronaut

Résultat attendu:

  • Image generated successfully
  • URL: https://cloud.inference.sh/outputs/abc123.png
  • Model: FLUX.2 Dev
  • Processing time: 3.2s

Utilisation de "agent-tools". Create a video of a drone flying over mountains

Résultat attendu:

  • Video task submitted
  • Task ID: xyz789
  • Estimated time: 45 seconds
  • Result: https://cloud.inference.sh/videos/xyz789.mp4

Audit de sécurité

Risque faible
v1 • 2/23/2026

Static analysis flagged 194 patterns, but most are false positives from documentation files. External command detections are markdown code blocks showing CLI examples, not executable code. Network URLs are documentation links. Environment variable access (INFSH_API_KEY) and filesystem paths (~/.local/bin) are standard for CLI tools. The curl|sh install pattern is common for CLI distribution. One medium concern: CLI executes external processes via inference.sh API, requiring user trust in the platform.

5
Fichiers analysés
554
Lignes analysées
4
résultats
1
Total des audits
Problèmes à risque moyen (1)
External CLI Dependency
Skill relies on external inference.sh CLI binary that executes cloud-based AI workloads. User must trust the inference.sh platform with API credentials and data sent for processing.
Problèmes à risque faible (1)
Standard CLI Installation Pattern
Uses curl|sh installation method which is common but requires trusting the remote script source.
Audité par: claude

Score de qualité

45
Architecture
100
Maintenabilité
87
Contenu
50
Communauté
83
Sécurité
83
Conformité aux spécifications

Ce que vous pouvez construire

AI Content Creation

Generate marketing images, product videos, and 3D models on-demand without local GPU. Create multiple variations quickly for A/B testing.

Rapid AI Prototyping

Test different AI models (FLUX vs Gemini, Veo vs Seedance) for your use case before committing to a specific provider or self-hosting.

Automated Workflows

Build automated pipelines that generate social media content, transcribe audio, or search the web as part of larger AI agent workflows.

Essayez ces prompts

Basic Image Generation
Generate an image using FLUX with this prompt: "a futuristic city at sunset, cyberpunk style"
Video Creation with Veo
Create a 5-second video using Veo 3.1 Fast showing "ocean waves crashing on rocky cliffs at sunrise"
Web Search Research
Search for the latest news about "quantum computing breakthroughs 2025" using Tavily and summarize the top 3 results
Multi-Model Comparison
Generate the same image prompt "a serene mountain landscape" using both FLUX and Gemini 3 Pro, then compare the results side by side

Bonnes pratiques

  • Use --save flag to save sample inputs as templates for repeated workflows
  • Check app details with infsh app get before running to understand required inputs
  • Use --no-wait for long-running video generation tasks to avoid blocking

Éviter

  • Do not send sensitive or private data to cloud AI models
  • Avoid running apps without checking input schema first
  • Do not store API keys in code - use environment variables

Foire aux questions

Do I need an inference.sh account to use this skill?
Yes. Run infsh login to authenticate. The skill requires valid API credentials to execute AI workloads.
Is my data private when using inference.sh?
All processing happens on inference.sh servers. Do not use for sensitive data. Check their privacy policy for data handling details.
How does pricing work?
inference.sh uses credits-based pricing. Different models cost different amounts per generation. Check your balance with infsh me.
Can I use this skill offline?
No. All AI inference happens in the cloud. An internet connection is required for all operations.
What happens if a task fails?
The CLI returns an error message with details. Failed tasks are not charged to your account.
Can I deploy my own AI models to inference.sh?
Yes. Use infsh app init and infsh app deploy to create and deploy custom AI workflows to the platform.

Détails du développeur

Structure de fichiers